[gentoo-project] call for agenda items, council meeting Mar 12

[gentoo-project] call for agenda items, council meeting Mar 12

[William Hubbs]

[-- Attachment #1: Type: text/plain, Size: 242 bytes --]

All,

the Gentoo council will meet again on March 12 at 19:00 utc in the
#gentoo-council channel on freenode.
Please respond to this message with any items you would like us to add
to the agenda to vote on or discuss.

Thanks much,

William


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 163 bytes --]

Re: [gentoo-project] call for agenda items, council meeting Mar 12

[Ulrich Mueller]

[-- Attachment #1: Type: text/plain, Size: 1314 bytes --]

>>>>> On Mon, 27 Feb 2017, William Hubbs wrote:

> the Gentoo council will meet again on March 12 at 19:00 utc in the
> #gentoo-council channel on freenode.
> Please respond to this message with any items you would like us to
> add to the agenda to vote on or discuss.

In the 2014-10-14 meeting, there was the following decision under the
"Git Migration Issues" topic:

   Can we drop CVS headers post-migration?

   Aye: blueness, creffett (proxy for ulm), dberkholz, dilfridge,
   radhermit, rich0, williamh 

This was again briefly discussed in the 2016-04-10 meeting (following
a discussion in the gentoo-dev mailing list) and in the 2016-11-13
meeting (with respect to a repoman check for the ebuild header).

Since there appear to be doubts how to interpret above mentioned
decision, I would like to ask the Council to clarify the following
points:

a) Are $Id$ and $Header$ lines to be removed in the gentoo repository?

b) Does this only apply to ebuilds and eclasses, or also to other
   files in the tree, e.g., metadata, profiles, and files in FILESDIR
   other than patches (like init scripts)?

c) Should these lines be removed in one go, or should we enable a
   repoman check and have them fade out over time?

d) Should git expansion of $Id$ be enabled (i.e., ident in git
   attributes)?

Ulrich

[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-project] call for agenda items, council meeting Mar 12

[Mike Gilbert]

On Mon, Feb 27, 2017 at 12:09 PM, Ulrich Mueller <ulm@gentoo.org> wrote:
>>>>>> On Mon, 27 Feb 2017, William Hubbs wrote:
>
>> the Gentoo council will meet again on March 12 at 19:00 utc in the
>> #gentoo-council channel on freenode.
>> Please respond to this message with any items you would like us to
>> add to the agenda to vote on or discuss.
>
> In the 2014-10-14 meeting, there was the following decision under the
> "Git Migration Issues" topic:
>
>    Can we drop CVS headers post-migration?
>
>    Aye: blueness, creffett (proxy for ulm), dberkholz, dilfridge,
>    radhermit, rich0, williamh
>
> This was again briefly discussed in the 2016-04-10 meeting (following
> a discussion in the gentoo-dev mailing list) and in the 2016-11-13
> meeting (with respect to a repoman check for the ebuild header).
>
> Since there appear to be doubts how to interpret above mentioned
> decision, I would like to ask the Council to clarify the following
> points:
>
> a) Are $Id$ and $Header$ lines to be removed in the gentoo repository?
>
> b) Does this only apply to ebuilds and eclasses, or also to other
>    files in the tree, e.g., metadata, profiles, and files in FILESDIR
>    other than patches (like init scripts)?
>
> c) Should these lines be removed in one go, or should we enable a
>    repoman check and have them fade out over time?
>
> d) Should git expansion of $Id$ be enabled (i.e., ident in git
>    attributes)?

If $Id$ is to be kept, I think point "d" needs further clarification:
in what contexts should ident expansion be enabled?

At rsync generation time?

In the development repo? (via .gitattributes).

Enabling expansion in some places but not others may cause some issues
that would necessitate further work to prevent "$Id: xxxxxxx $"
strings from being committed in the development repo accidentally.
This might mean a repoman check and/or a git hook.

I would highly suggest the council familiarize themselves with the git
ident attribute from the gitattributes man page before making any
decision here.

Re: [gentoo-project] call for agenda items, council meeting Mar 12

[Ulrich Mueller]

[-- Attachment #1: Type: text/plain, Size: 970 bytes --]

>>>>> On Mon, 27 Feb 2017, Mike Gilbert wrote:

>> d) Should git expansion of $Id$ be enabled (i.e., ident in git
>> attributes)?

> If $Id$ is to be kept, I think point "d" needs further clarification:
> in what contexts should ident expansion be enabled?

> At rsync generation time?

> In the development repo? (via .gitattributes).

> Enabling expansion in some places but not others may cause some issues
> that would necessitate further work to prevent "$Id: xxxxxxx $"
> strings from being committed in the development repo accidentally.
> This might mean a repoman check and/or a git hook.

Right, there are many paths to insanity there. That's why we should get
rid of the whole thing altogether.

A content tracker should faithfully keep track of the files committed,
but not modify their contents. Even CVS didn't reduce $Id: something$
to $Id$, but kept the original version in the repository [1].

Ulrich

[1] http://cvsman.com/cvs-1.12.12/cvs_99.php#SEC99

[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]

Re: [gentoo-project] call for agenda items, council meeting Mar 12

[William Hubbs]

[-- Attachment #1: Type: text/plain, Size: 505 bytes --]

On Mon, Feb 27, 2017 at 07:16:45PM +0100, Ulrich Mueller wrote:
> Right, there are many paths to insanity there. That's why we should get
> rid of the whole thing altogether.
 
 I am strongly with ulm on this, we should get rid of $Id$ expansion
 altogether, especially since there is a way to generate the hashes that
 $Id$ would create  from the command line.

 If folks want to use the hashes in a work flow some how they can, but
 they definitely do not belong in the ebuilds.

 William


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 163 bytes --]